Appendix C: Change History

C.1    Changes between 0.10 and 0.70:

• Removed references to the java.sql2 package. Substituted a proposal that splits the JDBC API into two packages: javax.sql and java.sql.
  
  
• Removed descriptions of the classes and intefaces in javax.sql. We plan to add them in again later, or create a separate document for the extended JDBC API.
  
  
• Removed the ScrollableResultSet interface and associated methods on other interfaces. Added methods for scrolling to the ResultSet interface
  
  
• Added result set type, concurrency type, keyset size, fetch size, and fetch direction properties to Connection, Statement, and ResultSet interfaces.
  
  
• Deprecated Statement.setCursorName and ResultSet.getCursorName. Use of "SELECT FOR UPDATE" is also deprecated. The new result set update methods can be used instead.
  
  
• Added SQL99 APIs.
  
  
• Added new metadata for persistent Java objects.
  
  

C.2    Changes between 0.70 and 0.80

• The Struct interface no longer extends the SQLData interface.
  
  
• All occurrences of the SQLType interface have been removed and replaced with 'String'.
  
  
• The Array.getArray() method now returns Object instead of Object[]. This allows an int[] array to be returned, for example.
  
  
• The specification now states that a Java programming language array may be passed as an input parameter via PreparedStatement.setObject().
  
  
• The semantics of ResultSet.isDeleted(), etc. is clarified.
  
  
• The DatabaseMetaData.getClass() method has been dropped due to the difficulties involved in loading classes directly from a database.
  
  
• The ResultSet.getBigDecimal() and CallableStatement.getBigDecimal() methods which take an `int scale' parameter have been deprecated.
  
  
• Statement.getResultSetType() now throws SQLException
  
  
• Added method Array.getBaseTypeName(). Array.getBaseType() now returns an int type code.
  
  
• DatabaseMetaData.getUDTs() now allows a fully qualified SQL name which may contain wildcard characters in the typeNamePattern parameter.
  
  

C.3    Changes between 0.80 and 0.90

• Section 9.4: Compiled code examples and removed syntax errors.
  
  
• Section 9.4.2: Added note on the independence of the Java programming language and SQL99 inheritance hierarchies.
  
  
• Section 6.1: Removed the requirement that indirect updates, such as those performed by a trigger that is fired, be included in the update counts returned by a batch update.
  
  
• Section 5.6: Clarified some technical details concerning inserting new rows in a result set.
  
  
• Section 7.3.1: Noted that the TYPE_NAME returned by getTypeInfo() may contain a vendor specific type name when the type code is JAVA_ODBECT. Previously, the TYPE_NAME had to be null.
  
  
• Removed method SQLInput.readStruct() and SQLOutput.writeStruct(). Read/ writeObject() should be used instead.
  
  
• Chapter 5: Revised and simplified the scrollable result set model.
  
  
• Section 11.3: Added note on character conversion.
  
  
• Section 10.3: Added description of time zone support.
  
  

C.4    Changes between 0.90 and 0.91

• Section 5.5: Added rules for selecting a supported result set type and concurrency type.
  
  
• Section 5.6: Added method ResultSet.cancelRowUpdates.
  
  
• Section 5.6: Added more description on the semantics of the update methods.
  
  
• Added Section 5.9.
  
  

C.5    Changes between 0.91 and 0.95

• Section 5.6: Added general description of the queries that yield updatable result sets.
  
  
• Section 8.2: Clarified rules for passing blob and clob values as input parameters.
  
  
• Sections 11.4 and 11.5: Added these sections to the document.
  
  

C.6    Changes between 0.95 and 1.0

• no changes
  
  

C.7    Changes between 1.0 and 1.1

• Chapter 6: Extended the semantics of executeBatch. Batch processing is now allowed to continue after an element in a batch has failed to execute successfully.
  
  
• Added Section 10.6.